issue_comments

For anyone else who runs into this, I built a minimal conda package for xarray 0.7 for Python 3.3. It's in the richardotis channel on anaconda.org. Since I make no promises it will be indefinitely maintained, here is the conda.yaml. Put the file in a directory by itself, cd into it and use conda build . --python 3.3 to build.

``` yaml package: name: xarray version: 0.7.0 source: git_url: https://github.com/pydata/xarray.git git_tag: v0.7.0 requirements: build: - python - setuptools - pandas run: - python - pandas

build: script: python setup.py install test: imports: - xarray about: home: https://github.com/pydata/xarray license: Apache summary: 'N-Dimensional Labeled Arrays' ```

That's what I'm going to do. Thanks for the help.

but with xarray being a pure Python package and seemingly happily running on the latest py33 versions of its dependencies

It would probably be just as easy for you to build xarray from source then. Or use pip. Both options should work with 3.3.

I don't think it's too much to ask for one more py33 build after the rename.

Yes and no. If you want the full optional dependencies, xarray can be a little bit hard to build (with pynio support, etc). If you want just the basics then it is relatively easy.

You should make the conda package request in https://github.com/ContinuumIO/anaconda-issues/issues/635, and other continuum communication channels (mailing list, gitter, etc).

It occurs to me that, since xarray is in the "defaults" channel, this probably has to do with a policy at Continuum and so I might be better off just hosting a py33 version of xarray in my own conda channel.

That's true. I just looked and saw numpy's last Python 3.3 build was 1.9.2. On the other hand, I feel that the xarray package rename has created a special case. Without the rename, pycalphad could continue its py33 support indefinitely as long as nothing happened to the relevant xarray APIs (not guaranteed, I understand that).

I'm happy to drop support for deprecated versions when I have to, but with xarray being a pure Python package and seemingly happily running on the latest py33 versions of its dependencies, I don't think it's too much to ask for one more py33 build after the rename.

The use of "official release" can be confusing in this case. I believe that the latest xarray works OK in Python 3.3. But note that the project is not tested on Python 3.3 since PR https://github.com/pydata/xarray/pull/583. I believe that the official PyPI source dist will install OK in Python 3.3 BTW.

On the other hand, Continuum does seem to be abandoning Python 3.3, there are hardly new conda packages being built for it.

So it appears that xray hasn't pushed a py33 conda release since 0.5.2, and I didn't notice because conda happily installed that version in my py33 tests and everything worked. With the name change, it can't find a compatible version anymore.

For dependency resolution purposes I would appreciate an official py33 release of xarray 0.7, or even "xarray 0.5.2" if there's been some kind of breaking change just so that conda can find the package.

I run py35 personally but my library pycalphad supports py27 and py33+. I noticed this issue when I updated my CI configuration to use "xarray" and only the py33 builds were failing.

@richardotis any special reason for wanting a Python 3.3 version? The migration to >=3.4 is highly recommended. (And I don't think continuum and the community building stuff for Python 3.3 lately.)

If you cannot upgrade you can use this recipe to build you own Python 3.3 xarray.

If you clone the whole repo you can then type:

shell export CONDA_PY=33 export CONDA_NPY=110 # (or any other numpy version series you need 18, 19...) conda build xarray

If you issue the last command from the repo root directory conda will build dependencies for you.